International Journal of Neuroscience
ISSN: 0020-7454 (Print) 1543-5245 (Online) Journal homepage: http://www.tandfonline.com/loi/ines20
Neuropsychological Impairment Among Intravenous Drug Users in Pre-Aids Stages of HIV Infection Mary Chelea Wellman To cite this article: Mary Chelea Wellman (1992) Neuropsychological Impairment Among Intravenous Drug Users in Pre-Aids Stages of HIV Infection, International Journal of Neuroscience, 64:1-4, 183-194, DOI: 10.3109/00207459209000544 To link to this article: http://dx.doi.org/10.3109/00207459209000544
Published online: 07 Jul 2009.
Submit your article to this journal
Article views: 2
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ines20 Download by: [University of California, San Diego]
Date: 17 March 2016, At: 00:05
0 1992 Gordon and Breach Science Publishers S.A.
Intern. J . Neuroscience, 1992, Vol. 64, pp. 183-194 Reprints available directly from the publisher Photocopying permitted by license only
Printed in the United States of America
NEUROPSYCHOLOGICAL IMPAIRMENT AMONG INTRAVENOUS DRUG USERS IN PRE-AIDS STAGES OF HIV INFECTION MARY CHELEA WELLMAN
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
Department of Counseling and Educational Psychology, Rhode Island College, Providence, RI 02908 (Received Ocrober 6,1991) While much of the current literature concurs that neuropsychological decline does not occur among gay men in the early stages of HIV infection, there is no comparable body of evidence with regard to seropositive intravenous drug users (IVDU). In this study, 45 seropositive (CDC groups 2, 3, and 4a) IVDU in recovery and 55 seronegative IVDU in recovery were given a complete battery of neuropsychological tests measuring attention, language, visual-motor, memory, and conceptual skills, The groups were not significantly different in age, incidence of childhood and adult head injury, types of drugs used, length of use of cocaine, crack, amphetamines and hallucinogens, overdose history, and length of time in recovery. In addition, groups were statistically corrected for education level and length of heroin use. Results indicate that the seropositive participants scored significantly lower on measures of divided attention, visual short-term memory, graphomotor speed and accuracy, auditory language shortterm memory and abstract concept formation. Further analyses revealed that 18% of participants with Persistent Generalized Lymphadenopathy (CDC group 111) and 27.% of those with constitutional disease (CDC group IVa) were neuropsychologically impaired, as their performance was two standard deviations or more below the normative mean on two or more measures. These results are similar to the reported performance of gay men with full-blown AIDS in a number of studies. It is hypothesized that because of premorbid neurological insult, the toxic effects of drug abuse on brain tissue, and the immunosuppressive effects of the drugs, subcortical brain cells of IVDU are more vulnerable to the invasion of HIV, and neurological deterioration may occur at earlier stages of HIV Spectrum Disease in IVDU than in gay men. Keywords: AIDS, Drug users, neuropsychological impairment, HIV.
In the second decade of the AIDS epidemic, over 200,000 people in the United States have been diagnosed with full-blown AIDS, and over two million are estimated to be infected with the human immunodeficiency virus (HIV). HJV has been shown to be neurotropic and well as lymphotropic; it directly invades the cells of the central nervous system as well as the immune system. Navia and Price (1987) named the resulting disorder AIDS Dementia Complex (ADC) and characterized it as encephalopathy accompanied by dementia and motor dysfunction. Price, Navia and Cho (1986) reported that neuropsychological symptoms can be I express my appreciation to Rhode Island College and in particular to President John Nazarian for sabbatical support, to the Fred M. Roddy Foundation, the RI College Faculty Research Committee, and to the Worcester Jewish Federation for grant support. I thank CODAC Drug Treatment, Benjamin Rush Detoxification Center, Spectrum House, and AIDS Project Worcester for sharing their facilities, clients, and wisdom. I am indebted to Marybeth Shears Blanchette for test scoring, data coding and analysis as well as patience and fortitude. I express thanks to Carla White, Ph.D., for Rey scoring, and to Robert Wellman, Ph.D, for statistical assistance and literature suggestions. Finally, I express my heartfelt thanks to the 100 research participants who taught me much about courage, faith, and perseverance. Data presented at the annual convention of the American Psychological Association on August 19, 1991 in San Francisco, CA. 183
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
I84
M.C. WELLMAN
manifested before the onset of full-blown AIDS. Some recent controversy exists in the literature concerning neuropsychological deficits among seropositive gay men with no symptoms of disease. A number of studies (e.g., Goethe, Mitchell, Marshall, Brey, Cahill, Leger, Hoy, Baswell, 1989; 0110, Johnson, and Grafman, 1991; Tross, Price, Navia, Thaler, Gold, Hirsch, and Sidtis, 1988) have found that asymptomatic gay men have no neuropsychological deficits. In a longitudinal study of healthy HIV+ gay/bisexual men, Selnes, Miller, McArthur, Gordon, Munoz, Sheridan, Fox and Saah (1990) stated that there was no decline in neuropsychological test performance over an average of 13 months. Some other researchers disagree, reporting that asymptomatic gay men have significantly lower neuropsychological test scores than do seronegative men. Grant, Atkinson, Hesselink, Kennedy, Richman, Spector and McCutchan (1987) created a stir in the literature by reporting that 44% of asymptomatic gay men suffered from neuropsychological deficits. However, their criteria for impairment was a score of one standard deviation below the mean on two tests, or two standard deviations below the mean on one test. These are quite rigorous criteria, and may result in identification of a number of statistically false positive cases. More recently, Lunn, Skydsbjerg, Schulsinger, Parnas, Pedersen, and Mathiesen (1991) found that 35% of their asymptomatic gay men were cognitively impaired. However, their criterion for impairment was a score in the lowest quartile (25th percentile or lower, less than one standard deviation below the mean) on two or more tests. This criterion is quite stringent, and resulted in 1/5 of the normal controls being classified as cognitively impaired. Stern, Marder, Bell, Chen, Dooneief, Goldstein, Mindry, Richards, Sano, Williams, Gorman, Ehrhardt and Mayeux (1991) reported that asymptomatic gay men performed significantly worse than seronegatives on tests of memory, executive function, attention and abstract reasoning. Researchers agree that once HIV Spectrum Disease progresses and medical symptoms are present, neuropsychological dysfunction is evident. Van Gorp, Miller, Satz, and Visscher (1989) found that motor speed, visual short-term memory, cognitive flexibility and visual analysis were affected in gay men who were in the early stages of ADC and were symptomatic of AIDS. The wave of HIV infection has increased dramatically among the intravenous drug user (IVDU) population. There is evidence that IVDU are frequently neurologically compromised prior to the onset of HIV infection because of a multitude of factors, including prior head injury, and the effects of the chemical dependence itself (Lewis, 1989). Penk, Brown, Roberts, Dolan, Atkins and Robinowitz (1981) found that both heroin and polysubstance abusers fell two standard deviations below their expected performance, based on IQ score, when tested with the Benton Visual Memory Test. This neuropsychological instrument is used to assess brain damage, and measures visual-motor and short-term visual memory functioning. Strang and Gurling (1989) reported abnormalities in the neuropsychological tests and computerized tomography assessments of long-term heroin addicts. In addition, some drugs have been shown to have immunosuppressive effects which may accelerate the proliferation of HIV once the virus invades the body and nervous system. Donahoe and Falek (1988) reported that opiate addicts show signs of immunosuppression even when seronegative for HIV. Thyss, Bessi and Kretzschmar (1982) found that heroin addicts had fewer helper-T and suppressor-T cells than did a sample of the general population, and concluded that prolonged heroin addiction compromises or destroys cellular immunity.
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
185
Canabidoids found in marijuana have immunomodulary effects as well, particularly in terms of resistance to retroviruses (Friedman, Klein, Specter, Pross, Newton, Blanchard, and Widen, 1988). Cocaine also suppresses immune functioning (Klein, Newton, and Friedman, 1988). Levy, Janssen, Bush, and Rosenblum (1988) reported that IVDU with full-blown AIDS had a significantly higher incidence of neurological illness than did people with AIDS who engaged in other risk behaviors. Thus, chronic use of various drugs has been found to be related to neurological dysfunction and also immunosuppression. As a result, it is often difficult to discern clinically whether neuropsychological dysfunction among HIV-infected IVDU is partially or fully caused by pre-existing conditions. Few systematic carefully designed studies using only IVDU as subjects have been done to examine the relationship of early stages of HIV infection and neuropsychological deficits in this population. Silberstein, McKegney , O’Dowd, Selwyn, Schoenbaum, Drucker, Feiner, Cox and Friedland ( 1987) studied nonmatched seronegative and seropositive IVDU “with no overt symptoms of AIDS-related illness” and administered eight neuropsychological measures. Univariate analysis differences were noted on finger tapping (dominant hand), Trailmaking A, and WAIS-R Similarities and Digit Span. The investigators reported that motor speed, concept formation, and associative learning were significantly more impaired in the seropositive subjects. In multivariate analyses, a difference was seen only in associative learning tasks. However, it is not known what stage of pre-AIDS (CDC 11, I11 or IVa) was present in each of the seropositive subjects in the Silberstein et al. study, as well as how many subjects with AIDS who had no acute illness at the time of evaluation were included. Further, the groups were not equated for education, IQ, or history of head injury, and were equated only for patterns of drug use in the multivariate analyses. METHOD Forty-five seropositive pre-AIDS intravenous drug users in recovery and fifty-five seronegative drug users in recovery from four drug and AIDS community agencies in New England consented to participate. They were recruited by their counselors at the community agency from which they received services. All gave written informed consent and were given a code number to protect their anonymity. All participants were between 18 and 45 years of age (half were between 25 and 34). Fifty-two of the participants were male, while 48 were female. Fourteen identified themselves as Latino/Latina, 2 1 identified themselves as African-American, and 65 identified themselves as Caucasian. Seropositives and seronegatives did not differ significantly in age, history of childhood and adult head injury, types of drugs used, length of use of cocaine, crack, and amphetamines, overdose history, and length of time in recovery (all p s > .30). Fifty-four percent of the seropositives and 46% of the seronegatives reported incidence of childhood head injury, while 55% of seropositives and 45% of seronegatives reported incidence of adult head injury, indicating that head injury is a fairly common characteristic in this sample. Seropositives and seronegatives also did not differ significantly in length of use of hallucinogens, involvements in methadone treatment, or length of time in methadone treatment (all p s > .lo). Alcohol use was not measured, a limitation of the study. No toxicological screenings were conducted at the time of the evaluations. Seropositives had significantly lower education level and significantly longer heroin use than did seronegatives. Differences in education level and length of heroin use
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
186
M.C. WELLMAN
were statistically controlled in data analyses, using analysis of covariance with education and heroin use as covariates. Seropositive participants were not selected based on CDC group, but interview data indicated that there were 8 asymptomatic (CDC Group II), 11 with Persistent Generalized Lymphadeneopathy (CDC Group 111), and 26 with constitutional disease (CDC Group IVa). Each evaluation session lasted about 90 minutes; participants were encouraged to take a 15-minute rest period halfway through the session. If fatigue was observed by the evaluator in any evaluation session, the participant was given more frequent rest periods. This was rarely necessary, and few participants verbalized feelings of fatigue. To the contrary, most participants indicated enjoyment of the tasks, and the desire to continue working. All subjects were evaluated by the author, a psychologist licensed and certified in the states where evaluations were performed. At the end of each evaluation session, subjects received a supermarket gift certificate or an equivalent gift package of a cash voucher and two books of the subject’s choosing about AIDS and addiction. Evaluations included assessment of both neuropsychological and affective functioning. The affective measures and results are described elsewhere (Wellman, 1991). Commonly used neuropsychological instruments were chosen to measure language, memory and visual-motor functioning. The following instruments were used: I . Lafayette Grooved Pegboard (Pegboard dominant hand (DH) and Pegboard nondominant hand (NDH)) (Trites, 1987)-measures motor speed and dexterity for both dominant and nondominant hands. Scores reflect the number of seconds needed for task completion. Using Trites’ published norms for dominant and nondominant hands, 2 SD or more below the mean was used to reflect impairment. 2. Wechsler Memory Scale-Revised (Wechsler, 1987): Logical Memory Immediate and Delayed (Logical Mem I and 11)-measures auditory short-term and recent memory for brief narratives read to the subject. Scores are expressed in percentiles from national norms. Percentiles < 3 (2 SD) reflected impairment. 3 . Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981) Scaled scores were used for all subtests. Scaled scores < 5 (2 SD) reflected impairment. a. Digit Span-measures auditory short-term memory for series of numbers read to the subject; Digits Reversed additionally measures mental manipulation and sequential recoding of digits. b. Vocabulary-measures long-term memory and expressive language for word meanings. c. Block Design-measures visual analysis of abstract visual stimuli and visuoconstructive abilities. d. Digit Symbol-measures graphomotor speed and accuracy as well as visual short-term memory e. Similarities-measures logical categorical thinking, conceptualization. 4. Trailmaking A and B (Trails A and Trails €3) (Reitan, 1979)-measures attention, attention with distraction, graphomotor speed, and sequential problem-solving. Scores reflect the number of seconds needed for task completion. Trails A scores of > 48 reflected impairment. Trails B scores of > 99 reflected impairment. 5. Rey-Osterrieth Complex Figure Test: copy, immediate recall, 30-minute delayed recall (Rey 1, Rey 2, Rey 3) (Rey, 1941) measures visuospatial memory and graphomotor visuoconstructive skill. Total raw scores were used for each administration. Scores corresponding to percentiles < 3 (2 SD) were judged to be impaired (Lezak, 1983).
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
187
TABLE 1 Neuropsychological Performance as a Function of HIV Status ~~
~
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mern I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
HIV + ( N = 45) X
HIV (N = 55) X
59.9 62.4 22.4 26.9 8.7 1.3 1.6 7.5 7.4 30.3 66.8 33.7 17.0 15.4
51.1 60.5 36.4 37.2 9.2 8.1 8.4 8.8 8.9 22.3 52.9 33.8 20.8 19.6
ANCOVA F value HIV
ED
.2 .I 8.4** 3.5
.2 1.5 4.1* 3.9* 4.8* 6.9** .I 4.2* 6.1* .6 2.3 5.6* .1 .I
.1
.7 2.8 3.9* 8.6** 12. I*** 5.9* .3 1.9 2.1
Heroin 0 2.4 8.2** 3.2 .1 2.8 .1 5.2* .3 1.3 .I 2.3 .I .2
df = 1 in all cases *p < .05 **p < .01 ***p < ,001
In addition, a structured interview inventory gathered information concerning demographic data, handedness, educational history, pediatric head injuries, accidents, loss of consciousness, physical abuse, adult head injuries, vehicular accidents, loss of consciousness, length and categories of substance abuse, overdoses, history of needle use and sharing, length of time in recovery from substance use, use of methadone maintenance, health history including HIV constitutional symptoms, sexually transmitted diseases and hepatitis, HIV status, length of time seropositivity status known, sexual history and condom use. Both seronegatives and seropositives were asked to report any symptoms of constitutional disease, including chronic fatigue, loss of appetite, persistent dry cough, oral Candida, night sweats, chronic diarrhea, and persistent generalized lymphadenopathy. Because these variables may impact on neuropsychological functioning of IVDU, participants were questioned about each factor, creating a comprehensive research data base. RESULTS Performance on all neuropsychological measures for seronegative and seropositive groups was compared using two-way analyses of covariance (ANCOVA) with serostatus and sex as independent variables, the neuropsychological test scores as dependent variables, and education level and length of heroin use as covariates. As Table 1 indicates, seropositive and seronegative groups differed significantly on the WMS Logical Memory I, WAIS-R Similarities, Digit Symbol, and Trailmaking A and B. In subsequent sets of ANCOVAS using HIV and drug use, HIV and recovery, and HIV and sexually transmitted diseases as independent variables, the Rey Osterreith scores as dependent variables, and education and heroin use as covariates, the Rey-Osterreith immediate and delayed recall were also significantly different be-
M.C. WELLMAN
I88
TABLE 2 Neuropsychological Performance as a Function of Stage of HIV Infection
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST
Pegboard DH Pegboard NDH Logical Mem. I Logical Mem. I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
CDC I1 X
CDC 111 X
CDC IVa X
ANCOVA F value
57.3 58.9 33.9 38.0 10.3 8.0 8.0 8.5 8.3 26.2 @.I 34.0 18.4 17.9
58.6 64.7 23.2 24.2 8.0 7.4 7.2 7.6 7.8 34.2 59.7 31.5 14.2 13.5
62.2 64.0 19.2 25.6 8.6 7.1 1.6 7.2 7.0 30.3 70.7 34.6 17.7 15.5
I .o .5 .9 1.1
1.4 .1 .4 .2 .4 .7 .5 3.4* .8 .6
*I, < .05 d/’ = 2 in all cases
tween HIV groups (e.g., Rey 2: F = 7.2, df = 1, p < .01; Rey 3: F = 6.7, df = 1, p < .05). On the Trailmaking and Pegboard tests, mean scores reflect the time needed to complete each task; therefore, a lower score reflects more proficient performance. The significant and highly significant differences on all these measures indicate that the seronegatives outperformed the seropositives on measures of auditory language short-term memory and visual short-term memory, abstract concept formation, graphornotor speed and accuracy, and divided attention. There was no significant difference between the groups on WAIS-R Digit Span, Vocabulary of Block Design, Lafayette Grooved Pegboard, Logical Memory 11, or the direct copy administration of the Rey Osterrieth, suggesting that the groups did not differ on rote auditory short-term memory for numbers, general word meaning skills, visual analysis, visuoconstructive skills, speed of motoric manipulation of objects, or delayed auditory recall. There was no significant sex difference in any analysis, except for WAIS-R Block Design ( p < .05) and Digit symbol ( p < .0l), indicating that males and females did not differ on any of the neuropsychological measures, except those involving visuoconstruction and graphomotor speed. The finding that males outperform females in abstract visuoconstructive tasks such as Block Design has been replicated numerous times (e.g., Wellman and Allen, 1983). Subsequent ANCOVAs using the same covariates and the independent variables of length of time in recovery from drug abuse, use of methadone, use of AZT or other medication, and length of time since HIV diagnosis, yielded no significant differences on neuropsychological test performance as a function of recovery time, methadone, AZT or other medications, or time since diagnosis. The reported differences in HIV + versus HIV-serostatus groups remained. Table 2 shows a comparison of CDC groups in performance on each of the neuropsychologicai measures. ANCOVAs using the hierarchical approach and regression approach were used to correct for unequal cell size, while education level and heroin use remained the covariates. CDC groups did not differ in level of perfor-
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
189
TABLE 3 Frequency of Impaired' Test Performance as a Function of HIV Status
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mem I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
HIVf ( N
=
45)
0% (O)*
0% (0) 4% (2) 7% (3) 4% ( 2 ) 4% (2) 0% (0) 9% (4) 4% (2) 13% (6) 18% (8) 7% (3) 20% (9) 13% ( 6 )
HIV- ( N
=
55)
0% (0) 0% (0) 5% (3) 4% (2) 0% (0) 0% (0) 0% (0) 0% (0) 2% (1) 0% (0) 5% (3) 4% (2) 13% (7) 9% ( 5 )
'Performance of two SD below the mean or lower 2percentage (N)
mance on any measure except for the copying trial of the Rey-Osterreith Complex Figure, on which participants with PGL performed significantly worse than the other two groups. T-test comparisons of seronegatives and seropositive asymptomatics (CDC group 11) on all neuropsychological measures yielded no significant differences. The percentage of seropositives and seronegatives who scored in the impaired range for each measure was calculated. When normative means and standard deviations for a test were available, performance of two standard deviations below the mean or lower was considered to be impaired. Cut-off scores for each instrument are described in the method section. The results are shown in Table 3. The percentage of seropositives in each CDC group who scored in the impaired range for each measure was also calculated. These results are shown in Table 4. Finally, the percentages of neuropsychologically impaired participants in seropositive and seronegative groups, and also the three CDC groups were calculated. Impairment is defined as having the scores on two or more measures in the impaired range. The results are shown in Table 5. Chi-square analyses using Yates' correction indicated a highly significant difference in the frequency of impairment between seronegatives and seropositives (? = 22.4, df = 1, p < .001). Chi-square analyses of the frequency of impairment among the three CDC groups was significant (x2= 7.2, df = 2 , p < .05). DISCUSSION These analyses show that pre-AIDS seropositive IVDU in recovery appear to have neuropsychological deficits in the areas of divided attention, visual short-term memory, graphomotor speed and accuracy, auditory contextual language short-term memory, and abstract concept formation. The difficulties in graphomotor speed and accuracy are consistent with the common clinical finding of handwriting deterioration in people with HIV Spectrum Disease (e.g., Forstein, 1989). These deficits cannot be accounted for by age, sex, education, incidence of prior head injury, length of
M.C. WELLMAN
190
TABLE 4 Frequency of Impaired' Test Performance as a Function of Stage of Infection STAGE OF INFECTION
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mem I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
CDC I1 ( N
=
8)
CDC 111 (N
0%(0)2 O%(O) O%(O)
=
O%(O) O%(O) O%(O) O%(O) 9%(1) O%(O) O%(O) 9%( 1) O%(O) 18%(2) 9%(1) 18%(2) 369'44) 9%(1)
O%(O) O%(O) O%(O) O%(O) O%(O) O%(O) 13%(1) 25%(2) 13%(1) 25%(2) O%(O)
11)
CDC IVa ( N
=
26)
O%(O) O%(O) 8%(2) 12%(3) 4%(1) 8%(2) O%(O) 12%(3) 8%(2) 12%(3) 19%(5) O%(O) 12%(3) 19%(5)
'Performance of two SD below the mean or lower 'percentage (N)
drug use, types of drugs used, overdose history, length of time in recovery, use of methadone, and use of AZT or other medication. Van Gorp et al. (1989) reported strikingly similar findings in gay men with fullblown AIDS. These men had significantly worse performance in WAIS-R Digit Symbol, Similarities, and Block Design, Trails A and B, Rey Ostemeth Delayed Recall, and RAVLT Delayed Recognition than did seronegative men. It is interesting to note that in the van Gorp et al. study, only people with AIDS, rather than ARC, significantly differed from seronegatives, whereas in the present study seropositives in pre-AIDS groups significantly differed from seronegatives. Similarly, Tross, Price, Navia, Thaler, Gold, Hirsch, and Sidtis (1988) found that gay men in the early stages of AIDS performed significantly worse than did seronegative gay men on WAIS-R Digit Symbol and Block Design, grooved pegboard, TABLE 5 Neuropsychological Impairment* as a Function of HIV Status and Stage of Infection
HIV( N = 55) HIV + ( N = 45) CDC I1 ( N = 8) CDC 111
Impaired
Unimpaired
11% (6)'
89% (49)
22% (10)
78% (35)
13% (1)
87% (7)
18% (2)
82% (9)
27% (7)
73% (19)
(N = 11)
CDC 1Va (N = 26)
*Neuropsychological impairment is herein defined as performance in the impaired range (2 SD below the mean or lower) on two or more neuropsychological measures. 'percentage (N)
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
191
and Trailmaking A and B, while seronegatives and asymptomatic seropositives did not differ on these or any of the other 12 measures. Some studies report that both heroin and cocaine cause immunosuppression, and over 90% of the subjects in the present study used IV heroin and cocaine in combination. In addition, there was a high incidence of head injury among the IVDU in this study. Both of these factors may contribute to the incidence of neuropsychological dysfunction among the seronegative participants. However, seropositive and seronegatives did not differ significantly in these pre-existing factors, and the seropositives had a significantly greater incidence of impairment than did the seronegatives. Thus, it is possible that once IVDU become infected with HIV, subcortical structures in the brain, which have been implicated as HIV targets in early stages of infection, may become differentially more vulnerable to invading viruses at the time of initial acute infection (CDC Group I). In fact, recent studies by Daar, Moudgil, Meyer, and Ho (1991), and Clark, Saag, Decker, Campbell-Hill, Roberson, Veldkamp, Kappes, Hahn, and Shaw (1991) report that at the time of initial infection, a large explosion of virus replication occurs. The latter researchers found that one milliliter of blood from a newly infected individual may contain 1000 tissue-cultureinfective doses, a level that then drops dramatically when attacked by the body’s Tcells. The data in the present research suggest that there may be an initial “vulnerability factor” (van Gorp, 1991) caused by the pre-existing factors of history of head injury and chronic drug use which cause weakened subcortical structures to be invaded by HIV when the initial explosion of viral replication occurs at CDC stage I. The findings of the present study are also in contrast to studies of gay men which report that the percentage of impaired participants does not increase significantly until participants with AIDS are entered into the equation. For example, Selnes, Miller, Becker, Cohen, McArthur, Visscher, Gordon, Satz, Ginzburg and Polk (1988), using the MACS data, reported that when comparing the performance of 786 seronegtive gay men and 757 seropositive (CDC Groups I1 and 111) on eight neuropsychological measures, 13% of seronegatives and 15% of seropositives were impaired (two measures more than 2 SD below the mean). This difference was not significant. Likewise, Miller, Selnes, McArthur, Satz, Becker, Cohen, Sheridan, Machado, van Gorp, and Visscher (1990), also using the MACS data, reported that on ten neuropsychological measures, 769 seronegatives showed 3.9% impairment (same criteria for impairment as above) while 727 seropositives (CDC Groups I1 and 111) showed 5.5% impairment. While these percentages of impairment were substantially lower than those reported in the first paper, again the differences between seropositives and seronegatives was not significant. Tross et al. (1988) also reported no difference in percentages of impaired participants when comparing seropositives and seronegatives on 17 neuropsychological measures. They found 0% of both seropositives and seronegatives had two “abnormal” tests, while 20% of participants in the early stages of AIDS did have two abnormal tests. In the present study, 18% of IVDU participants with PGL (CDC group 111) had two or more abnormal tests, and 27% of those with constitutional disease (CDC group IVa) were similarly impaired. Finally, Wilkie, Eisendorfer, Morgan, Lowenstein, and Szapocznik ( I 990) reported in their narrative that when comparing 46 seropositive (CDC Groups I1 and 111) gay men with 13 seronegative gay men using 12 traditional neuropsychological measures, 0% of the seronegatives and 3% of the seropositives were impaired (two measures more than 2 SD below the mean), while their Table 2 shows 0% of both groups impaired. In either case, the differences were not significant, even though
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
I92
M.C. WELLMAN
the authors did find specific significant differences between seropositives and seronegatives in four of the twelve traditional neuropsychological measures. Thus, these data suggest that neuropsychological impairment appears earlier in the course of HIV infection among lVDU than it does among gay men. When comparing the scores of the seronegatives and the seropositive asymptomatics (CDC Group 11) on the neuropsychological measures, there were no significant differences in any of the measures. However, as infection progresses to stages of PGL and constitutional disease, impairment was noted in auditory short-term memory for contextual language, logical categorical thinking, graphomotor speed and visual short-term memory. Such deficits are seen frequently in gay men only at the stage of full-blown AlDS. It is hypothesized that this escalation of neuropsychological impairment in IVDU may occur as a result of the initial vulnerability of subcortical brain tissue in these IVDU. Gay men in a number of studies showed few if any signs of neuropsychological impairment at pre-A1DS stages, probably because the gay men in most studies had no history of head injury or drug use, and thus lacked the hypothesized vulnerability factor. Another important point needs to be made concerning the 11% rate of impairment among the seronegatives. While all seronegatives had received their last HIV negative test results less than 12 months prior to the study, two of the six impaired seronegative participants reported three symptoms of constitutional disease upon interview, and two of the six reported four symptoms of constitutional disease. Thus, it is possible that these four participants’ last HIV test failed to detect HIV infection and falsely indicated seronegativity. Grant and Heaton (1990) voiced their concerns about the validity of HIV test results when selecting subjects for research, stating that some people may test negative when in fact they are positive. It is common knowledge that false negatives exist, as the ELISA and Western Blot tests only identify the presence of antibodies, rather than HIV itself. Other research studies have not reported any investigation of possible symptomatology in their seronegative subjects, so this problem of false negatives used in research may be fairly widespread. In this study, if the four out of six impaired seronegative participants who reported three or four constitutional symptoms truly are seropositive, then the actual rate of impairment among seronegatives is 2/51, or 4%, rather than 6/55, or 11%. Thus, a comparison between the percentage of impaired lVDU seronegatives and seropositives, even the asymptomatic seropositives (CDC Group Il), would yield quite different results, and would provide even more strong support for the notion that HIV affects neuropsychological functioning in IVDU early in the course of infection.
REFERENCES Clark, S . , Saag, M . , Decker, W . , Campbell-Hill, S . , Roberson, J . , Veldkamp, P . , Kappes, J . , Hahn, B., & Shaw, G . (1991). High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-I infection. The New England Journal of Medicine, 324, 954-960. Daar, E., Moudgil, T., Meyer, R . , & Ho, D. (1991). Transient high levels of viremia in patients with primary human immunodeficiency virus type 1 infection. The New England Journal of Medicine, 324, 961-964. Donahoe, R. & Falek, A. (1988). Neurc)immunomodulation by opiates and other drugs of abuse: Relationship to HIV infection and AIDS. In T. P. Bridge (Ed.) Psychological, neuropsychiafric and substance abuse aspects of AIDS. New York: Raven Press. Forstein, M . (1989, October). Neurological aspects of HlV infection. Colloquium presented at University of Massachusetts Medical Center, Worcester, MA.
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
193
Goethe, K., Mitchell, J., Marshall, D., Brey, R., Cahill, W . , Leger, G., Hoy, L., & Baswell, R. (1989). Neuropsychological and neurological function of human immunodeficiency virus seropositive asymptomatic individuals. Archives of Neurology, 4 6 , 129- 133. Grant, I . , Atkinson, J., Hesselink, J., Kennedy, C., Richman, D., Spector, S., & McCutchan, J. (1987). Evidence of early central nervous system involvement in AIDS and other HIV infections. Annals of Internal Medicine, 107, 828-836. Grant, I . & Heaton, R. (1990). Human immunodeficiency virus-type 1 (HIV-I) and the brain. Journal of Consulting and Clinical Psychology, 58, 22-30. Levy, R., Janssen, R., Bush, T., & Rosenblum, M. (1988). Neuroepidemiology of acquired immunodeficiency syndrome. Journal of Acquired Immune Deficiency Syndromes, 1, 3 1-40, Lewis, B. Personal communication, August 17, 1989. Lezak, M. (1983). Neuropsychological assessment (2nd ed). New York Oxford University Press. Lunn, S., Skydsbjerg, M., Schulsinger, H., Pamas, J., Pedersen, C., & Mathiesen, L. (1991). A preliminary report on the neuropsychologic sequelae of human immunodeficiency virus. Archives of General Psychiatry, 4 8 , 139-142. Miller, E . , Selnes. O., McArthur, J., Satz, P., Becker, J . , Cohen, B., Sheridan, K., Machado, A , , van Gorp, W., & Visscher, B. (1990). Neuropsychological performance in HIV-1-infected homosexual men: The Multicenter AIDS Cohort Study. Neurology, 4 0 , 197-203. Navia, B. & Price, R. (1987). The AIDS Dementia Complex as the presenting or sole manifestation of HIV Infection. Archives of Neurology, 4 4 , 65-69. 0110, C., Johnson, R., & Grafman, J. (1991). Signs of cognitive change in HIV disease: An eventrelated brain potential study. Neurology, 4 1 , 209-215. Penk, W., Brown, A,, Roberts, W., Dolan, M., Atkins, H., & Robinowitz, R. (1981). Visual memory for Black and White male heroin and non-heroin drug users. Journal of Abnormal Psychology, 90, 486-489. Price, R.,Navia, B., & Cho, E. (1986). AIDS encephalopathy. Neurologic Clinics, 4 , 285-301. Reitan, R. (1986). Trail Making Test: Manual for administration and scoring. Tucson, AZ: Reitan Neuropsychological Laboratory. Rey, A. (1941). L’examen psychologique dans les cas d’encephalopathie traumatique. Archives de Psychologie, 28, 286-340. Selnes, O., Miller, E., Becker, J., Cohen, B., McArthur, J . , Visscher, B., Gordon, B., Satz, P., Ginzburg, H., Polk, B. (1988, June). Normal neuropsychological performance in healthy HIV-1 infected men: The Multicenter AIDS Cohort Study (MACS). Paper presented at the IVth International Conference on AIDS, Stockholm, Sweden, June 15-16, 1988. Selnes, O., Miller, E., McArthur, J., Gordon, B., Munoz, A , , Sheridan, K., Fox, R., & Saah, A. (1990). HIV-1 infection: No evidence of cognitive decline during the asymptomatic stages. Neurology, 4 0 , 204-208. Silberstein, C., McKegney, F. P., O’Dowd, M., Selwyn, P., Schoenbaum, E., Drucker, E., Feiner, C . , Cox, C., Friedland, G. (1987). A prospective longitudinal study of neuropsychological and psychosocial factors in asymptomatic individuals at nsk for HTLV-III/LAV infection in a methadone program: Preliminary findings. International Journal of Neuroscience, 32, 669-676. Stem, Y., Marder, K., Bell, K., Chen, J., Dooneief, G., Goldstein, S., Mindry, D., Richards, M., Sano, M., Williams, J., Gorman, J . , Ehrhardt, A , , Mayeux, R. (1991). Multidisciplinary baseline assessment of homosexual men with and without human immunodeficiency virus: 111. Neurologic and neuropsychological findings. Archives of General Psychiatry, 4 8 . 13 1 - 138. Strang, J . & Gurling, H. (1989). Computerized tomography and neuropsychological assessment in longterm high dose heroin addicts. British Journal of the Addictions, 84. 1011-1019. Thyss, A,, Bessi, C., & Kretzschmar, J . (1982). Cellular immunity in heroin addicts. Psychologie Medicale, 14, 1731-1733. Trites, R. (1987). Instructions for the 32025 Grooved Pegboard Test. Lafayette, IN: Lafayette Instrument Company. Tross, S., Price, R., Navia, B., Thaler, H., Gold, J., Hirsch, D., & Sidtis, I. (1988). Neuropsychological characterization of the AIDS Dementia Complex: A preliminary report. AIDS,2 . 81-88. van Gorp, W. Personal communication, January 24, 1991. van Gorp, W., Miller, E., Satz, P., & Visscher, B. (1989). Neuropsychological performance in HIV1 immunocompromised patients: A preliminary report. Journal of Clinical and Experimental Neuropsychology, 11, 763-773. Wechsler, D. (198 1). Wechsler Adult Intelligence Scale-Revised. New York: The Psychological Corporation. Wechsler, D. (1987). Wechsler Memory Scale-Revised. San Antonio, TX: The Psychological Corporation.
I94
M.C. WELLMAN
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
Wellman, M. C . (1991). Psychological distress among intravenous drug users with HIV infection. Manuscript submitted for publication. Wellman, M . & Allen, M . (1983). Variations in hand position, cerebral lateralization, and reading ability among right-handed children. Brain and Language. 18, 217-292. Wilkie, F . , Eisdorfer, C . , Morgan, R . , Lowenstein, D . , Szapocznik, J . (1990). Cognition in early human immunodeficiency virus infection. Archives of Neurology, 47, 433-440.
ISSN: 0020-7454 (Print) 1543-5245 (Online) Journal homepage: http://www.tandfonline.com/loi/ines20
Neuropsychological Impairment Among Intravenous Drug Users in Pre-Aids Stages of HIV Infection Mary Chelea Wellman To cite this article: Mary Chelea Wellman (1992) Neuropsychological Impairment Among Intravenous Drug Users in Pre-Aids Stages of HIV Infection, International Journal of Neuroscience, 64:1-4, 183-194, DOI: 10.3109/00207459209000544 To link to this article: http://dx.doi.org/10.3109/00207459209000544
Published online: 07 Jul 2009.
Submit your article to this journal
Article views: 2
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ines20 Download by: [University of California, San Diego]
Date: 17 March 2016, At: 00:05
0 1992 Gordon and Breach Science Publishers S.A.
Intern. J . Neuroscience, 1992, Vol. 64, pp. 183-194 Reprints available directly from the publisher Photocopying permitted by license only
Printed in the United States of America
NEUROPSYCHOLOGICAL IMPAIRMENT AMONG INTRAVENOUS DRUG USERS IN PRE-AIDS STAGES OF HIV INFECTION MARY CHELEA WELLMAN
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
Department of Counseling and Educational Psychology, Rhode Island College, Providence, RI 02908 (Received Ocrober 6,1991) While much of the current literature concurs that neuropsychological decline does not occur among gay men in the early stages of HIV infection, there is no comparable body of evidence with regard to seropositive intravenous drug users (IVDU). In this study, 45 seropositive (CDC groups 2, 3, and 4a) IVDU in recovery and 55 seronegative IVDU in recovery were given a complete battery of neuropsychological tests measuring attention, language, visual-motor, memory, and conceptual skills, The groups were not significantly different in age, incidence of childhood and adult head injury, types of drugs used, length of use of cocaine, crack, amphetamines and hallucinogens, overdose history, and length of time in recovery. In addition, groups were statistically corrected for education level and length of heroin use. Results indicate that the seropositive participants scored significantly lower on measures of divided attention, visual short-term memory, graphomotor speed and accuracy, auditory language shortterm memory and abstract concept formation. Further analyses revealed that 18% of participants with Persistent Generalized Lymphadenopathy (CDC group 111) and 27.% of those with constitutional disease (CDC group IVa) were neuropsychologically impaired, as their performance was two standard deviations or more below the normative mean on two or more measures. These results are similar to the reported performance of gay men with full-blown AIDS in a number of studies. It is hypothesized that because of premorbid neurological insult, the toxic effects of drug abuse on brain tissue, and the immunosuppressive effects of the drugs, subcortical brain cells of IVDU are more vulnerable to the invasion of HIV, and neurological deterioration may occur at earlier stages of HIV Spectrum Disease in IVDU than in gay men. Keywords: AIDS, Drug users, neuropsychological impairment, HIV.
In the second decade of the AIDS epidemic, over 200,000 people in the United States have been diagnosed with full-blown AIDS, and over two million are estimated to be infected with the human immunodeficiency virus (HIV). HJV has been shown to be neurotropic and well as lymphotropic; it directly invades the cells of the central nervous system as well as the immune system. Navia and Price (1987) named the resulting disorder AIDS Dementia Complex (ADC) and characterized it as encephalopathy accompanied by dementia and motor dysfunction. Price, Navia and Cho (1986) reported that neuropsychological symptoms can be I express my appreciation to Rhode Island College and in particular to President John Nazarian for sabbatical support, to the Fred M. Roddy Foundation, the RI College Faculty Research Committee, and to the Worcester Jewish Federation for grant support. I thank CODAC Drug Treatment, Benjamin Rush Detoxification Center, Spectrum House, and AIDS Project Worcester for sharing their facilities, clients, and wisdom. I am indebted to Marybeth Shears Blanchette for test scoring, data coding and analysis as well as patience and fortitude. I express thanks to Carla White, Ph.D., for Rey scoring, and to Robert Wellman, Ph.D, for statistical assistance and literature suggestions. Finally, I express my heartfelt thanks to the 100 research participants who taught me much about courage, faith, and perseverance. Data presented at the annual convention of the American Psychological Association on August 19, 1991 in San Francisco, CA. 183
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
I84
M.C. WELLMAN
manifested before the onset of full-blown AIDS. Some recent controversy exists in the literature concerning neuropsychological deficits among seropositive gay men with no symptoms of disease. A number of studies (e.g., Goethe, Mitchell, Marshall, Brey, Cahill, Leger, Hoy, Baswell, 1989; 0110, Johnson, and Grafman, 1991; Tross, Price, Navia, Thaler, Gold, Hirsch, and Sidtis, 1988) have found that asymptomatic gay men have no neuropsychological deficits. In a longitudinal study of healthy HIV+ gay/bisexual men, Selnes, Miller, McArthur, Gordon, Munoz, Sheridan, Fox and Saah (1990) stated that there was no decline in neuropsychological test performance over an average of 13 months. Some other researchers disagree, reporting that asymptomatic gay men have significantly lower neuropsychological test scores than do seronegative men. Grant, Atkinson, Hesselink, Kennedy, Richman, Spector and McCutchan (1987) created a stir in the literature by reporting that 44% of asymptomatic gay men suffered from neuropsychological deficits. However, their criteria for impairment was a score of one standard deviation below the mean on two tests, or two standard deviations below the mean on one test. These are quite rigorous criteria, and may result in identification of a number of statistically false positive cases. More recently, Lunn, Skydsbjerg, Schulsinger, Parnas, Pedersen, and Mathiesen (1991) found that 35% of their asymptomatic gay men were cognitively impaired. However, their criterion for impairment was a score in the lowest quartile (25th percentile or lower, less than one standard deviation below the mean) on two or more tests. This criterion is quite stringent, and resulted in 1/5 of the normal controls being classified as cognitively impaired. Stern, Marder, Bell, Chen, Dooneief, Goldstein, Mindry, Richards, Sano, Williams, Gorman, Ehrhardt and Mayeux (1991) reported that asymptomatic gay men performed significantly worse than seronegatives on tests of memory, executive function, attention and abstract reasoning. Researchers agree that once HIV Spectrum Disease progresses and medical symptoms are present, neuropsychological dysfunction is evident. Van Gorp, Miller, Satz, and Visscher (1989) found that motor speed, visual short-term memory, cognitive flexibility and visual analysis were affected in gay men who were in the early stages of ADC and were symptomatic of AIDS. The wave of HIV infection has increased dramatically among the intravenous drug user (IVDU) population. There is evidence that IVDU are frequently neurologically compromised prior to the onset of HIV infection because of a multitude of factors, including prior head injury, and the effects of the chemical dependence itself (Lewis, 1989). Penk, Brown, Roberts, Dolan, Atkins and Robinowitz (1981) found that both heroin and polysubstance abusers fell two standard deviations below their expected performance, based on IQ score, when tested with the Benton Visual Memory Test. This neuropsychological instrument is used to assess brain damage, and measures visual-motor and short-term visual memory functioning. Strang and Gurling (1989) reported abnormalities in the neuropsychological tests and computerized tomography assessments of long-term heroin addicts. In addition, some drugs have been shown to have immunosuppressive effects which may accelerate the proliferation of HIV once the virus invades the body and nervous system. Donahoe and Falek (1988) reported that opiate addicts show signs of immunosuppression even when seronegative for HIV. Thyss, Bessi and Kretzschmar (1982) found that heroin addicts had fewer helper-T and suppressor-T cells than did a sample of the general population, and concluded that prolonged heroin addiction compromises or destroys cellular immunity.
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
185
Canabidoids found in marijuana have immunomodulary effects as well, particularly in terms of resistance to retroviruses (Friedman, Klein, Specter, Pross, Newton, Blanchard, and Widen, 1988). Cocaine also suppresses immune functioning (Klein, Newton, and Friedman, 1988). Levy, Janssen, Bush, and Rosenblum (1988) reported that IVDU with full-blown AIDS had a significantly higher incidence of neurological illness than did people with AIDS who engaged in other risk behaviors. Thus, chronic use of various drugs has been found to be related to neurological dysfunction and also immunosuppression. As a result, it is often difficult to discern clinically whether neuropsychological dysfunction among HIV-infected IVDU is partially or fully caused by pre-existing conditions. Few systematic carefully designed studies using only IVDU as subjects have been done to examine the relationship of early stages of HIV infection and neuropsychological deficits in this population. Silberstein, McKegney , O’Dowd, Selwyn, Schoenbaum, Drucker, Feiner, Cox and Friedland ( 1987) studied nonmatched seronegative and seropositive IVDU “with no overt symptoms of AIDS-related illness” and administered eight neuropsychological measures. Univariate analysis differences were noted on finger tapping (dominant hand), Trailmaking A, and WAIS-R Similarities and Digit Span. The investigators reported that motor speed, concept formation, and associative learning were significantly more impaired in the seropositive subjects. In multivariate analyses, a difference was seen only in associative learning tasks. However, it is not known what stage of pre-AIDS (CDC 11, I11 or IVa) was present in each of the seropositive subjects in the Silberstein et al. study, as well as how many subjects with AIDS who had no acute illness at the time of evaluation were included. Further, the groups were not equated for education, IQ, or history of head injury, and were equated only for patterns of drug use in the multivariate analyses. METHOD Forty-five seropositive pre-AIDS intravenous drug users in recovery and fifty-five seronegative drug users in recovery from four drug and AIDS community agencies in New England consented to participate. They were recruited by their counselors at the community agency from which they received services. All gave written informed consent and were given a code number to protect their anonymity. All participants were between 18 and 45 years of age (half were between 25 and 34). Fifty-two of the participants were male, while 48 were female. Fourteen identified themselves as Latino/Latina, 2 1 identified themselves as African-American, and 65 identified themselves as Caucasian. Seropositives and seronegatives did not differ significantly in age, history of childhood and adult head injury, types of drugs used, length of use of cocaine, crack, and amphetamines, overdose history, and length of time in recovery (all p s > .30). Fifty-four percent of the seropositives and 46% of the seronegatives reported incidence of childhood head injury, while 55% of seropositives and 45% of seronegatives reported incidence of adult head injury, indicating that head injury is a fairly common characteristic in this sample. Seropositives and seronegatives also did not differ significantly in length of use of hallucinogens, involvements in methadone treatment, or length of time in methadone treatment (all p s > .lo). Alcohol use was not measured, a limitation of the study. No toxicological screenings were conducted at the time of the evaluations. Seropositives had significantly lower education level and significantly longer heroin use than did seronegatives. Differences in education level and length of heroin use
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
186
M.C. WELLMAN
were statistically controlled in data analyses, using analysis of covariance with education and heroin use as covariates. Seropositive participants were not selected based on CDC group, but interview data indicated that there were 8 asymptomatic (CDC Group II), 11 with Persistent Generalized Lymphadeneopathy (CDC Group 111), and 26 with constitutional disease (CDC Group IVa). Each evaluation session lasted about 90 minutes; participants were encouraged to take a 15-minute rest period halfway through the session. If fatigue was observed by the evaluator in any evaluation session, the participant was given more frequent rest periods. This was rarely necessary, and few participants verbalized feelings of fatigue. To the contrary, most participants indicated enjoyment of the tasks, and the desire to continue working. All subjects were evaluated by the author, a psychologist licensed and certified in the states where evaluations were performed. At the end of each evaluation session, subjects received a supermarket gift certificate or an equivalent gift package of a cash voucher and two books of the subject’s choosing about AIDS and addiction. Evaluations included assessment of both neuropsychological and affective functioning. The affective measures and results are described elsewhere (Wellman, 1991). Commonly used neuropsychological instruments were chosen to measure language, memory and visual-motor functioning. The following instruments were used: I . Lafayette Grooved Pegboard (Pegboard dominant hand (DH) and Pegboard nondominant hand (NDH)) (Trites, 1987)-measures motor speed and dexterity for both dominant and nondominant hands. Scores reflect the number of seconds needed for task completion. Using Trites’ published norms for dominant and nondominant hands, 2 SD or more below the mean was used to reflect impairment. 2. Wechsler Memory Scale-Revised (Wechsler, 1987): Logical Memory Immediate and Delayed (Logical Mem I and 11)-measures auditory short-term and recent memory for brief narratives read to the subject. Scores are expressed in percentiles from national norms. Percentiles < 3 (2 SD) reflected impairment. 3 . Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981) Scaled scores were used for all subtests. Scaled scores < 5 (2 SD) reflected impairment. a. Digit Span-measures auditory short-term memory for series of numbers read to the subject; Digits Reversed additionally measures mental manipulation and sequential recoding of digits. b. Vocabulary-measures long-term memory and expressive language for word meanings. c. Block Design-measures visual analysis of abstract visual stimuli and visuoconstructive abilities. d. Digit Symbol-measures graphomotor speed and accuracy as well as visual short-term memory e. Similarities-measures logical categorical thinking, conceptualization. 4. Trailmaking A and B (Trails A and Trails €3) (Reitan, 1979)-measures attention, attention with distraction, graphomotor speed, and sequential problem-solving. Scores reflect the number of seconds needed for task completion. Trails A scores of > 48 reflected impairment. Trails B scores of > 99 reflected impairment. 5. Rey-Osterrieth Complex Figure Test: copy, immediate recall, 30-minute delayed recall (Rey 1, Rey 2, Rey 3) (Rey, 1941) measures visuospatial memory and graphomotor visuoconstructive skill. Total raw scores were used for each administration. Scores corresponding to percentiles < 3 (2 SD) were judged to be impaired (Lezak, 1983).
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
187
TABLE 1 Neuropsychological Performance as a Function of HIV Status ~~
~
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mern I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
HIV + ( N = 45) X
HIV (N = 55) X
59.9 62.4 22.4 26.9 8.7 1.3 1.6 7.5 7.4 30.3 66.8 33.7 17.0 15.4
51.1 60.5 36.4 37.2 9.2 8.1 8.4 8.8 8.9 22.3 52.9 33.8 20.8 19.6
ANCOVA F value HIV
ED
.2 .I 8.4** 3.5
.2 1.5 4.1* 3.9* 4.8* 6.9** .I 4.2* 6.1* .6 2.3 5.6* .1 .I
.1
.7 2.8 3.9* 8.6** 12. I*** 5.9* .3 1.9 2.1
Heroin 0 2.4 8.2** 3.2 .1 2.8 .1 5.2* .3 1.3 .I 2.3 .I .2
df = 1 in all cases *p < .05 **p < .01 ***p < ,001
In addition, a structured interview inventory gathered information concerning demographic data, handedness, educational history, pediatric head injuries, accidents, loss of consciousness, physical abuse, adult head injuries, vehicular accidents, loss of consciousness, length and categories of substance abuse, overdoses, history of needle use and sharing, length of time in recovery from substance use, use of methadone maintenance, health history including HIV constitutional symptoms, sexually transmitted diseases and hepatitis, HIV status, length of time seropositivity status known, sexual history and condom use. Both seronegatives and seropositives were asked to report any symptoms of constitutional disease, including chronic fatigue, loss of appetite, persistent dry cough, oral Candida, night sweats, chronic diarrhea, and persistent generalized lymphadenopathy. Because these variables may impact on neuropsychological functioning of IVDU, participants were questioned about each factor, creating a comprehensive research data base. RESULTS Performance on all neuropsychological measures for seronegative and seropositive groups was compared using two-way analyses of covariance (ANCOVA) with serostatus and sex as independent variables, the neuropsychological test scores as dependent variables, and education level and length of heroin use as covariates. As Table 1 indicates, seropositive and seronegative groups differed significantly on the WMS Logical Memory I, WAIS-R Similarities, Digit Symbol, and Trailmaking A and B. In subsequent sets of ANCOVAS using HIV and drug use, HIV and recovery, and HIV and sexually transmitted diseases as independent variables, the Rey Osterreith scores as dependent variables, and education and heroin use as covariates, the Rey-Osterreith immediate and delayed recall were also significantly different be-
M.C. WELLMAN
I88
TABLE 2 Neuropsychological Performance as a Function of Stage of HIV Infection
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST
Pegboard DH Pegboard NDH Logical Mem. I Logical Mem. I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
CDC I1 X
CDC 111 X
CDC IVa X
ANCOVA F value
57.3 58.9 33.9 38.0 10.3 8.0 8.0 8.5 8.3 26.2 @.I 34.0 18.4 17.9
58.6 64.7 23.2 24.2 8.0 7.4 7.2 7.6 7.8 34.2 59.7 31.5 14.2 13.5
62.2 64.0 19.2 25.6 8.6 7.1 1.6 7.2 7.0 30.3 70.7 34.6 17.7 15.5
I .o .5 .9 1.1
1.4 .1 .4 .2 .4 .7 .5 3.4* .8 .6
*I, < .05 d/’ = 2 in all cases
tween HIV groups (e.g., Rey 2: F = 7.2, df = 1, p < .01; Rey 3: F = 6.7, df = 1, p < .05). On the Trailmaking and Pegboard tests, mean scores reflect the time needed to complete each task; therefore, a lower score reflects more proficient performance. The significant and highly significant differences on all these measures indicate that the seronegatives outperformed the seropositives on measures of auditory language short-term memory and visual short-term memory, abstract concept formation, graphornotor speed and accuracy, and divided attention. There was no significant difference between the groups on WAIS-R Digit Span, Vocabulary of Block Design, Lafayette Grooved Pegboard, Logical Memory 11, or the direct copy administration of the Rey Osterrieth, suggesting that the groups did not differ on rote auditory short-term memory for numbers, general word meaning skills, visual analysis, visuoconstructive skills, speed of motoric manipulation of objects, or delayed auditory recall. There was no significant sex difference in any analysis, except for WAIS-R Block Design ( p < .05) and Digit symbol ( p < .0l), indicating that males and females did not differ on any of the neuropsychological measures, except those involving visuoconstruction and graphomotor speed. The finding that males outperform females in abstract visuoconstructive tasks such as Block Design has been replicated numerous times (e.g., Wellman and Allen, 1983). Subsequent ANCOVAs using the same covariates and the independent variables of length of time in recovery from drug abuse, use of methadone, use of AZT or other medication, and length of time since HIV diagnosis, yielded no significant differences on neuropsychological test performance as a function of recovery time, methadone, AZT or other medications, or time since diagnosis. The reported differences in HIV + versus HIV-serostatus groups remained. Table 2 shows a comparison of CDC groups in performance on each of the neuropsychologicai measures. ANCOVAs using the hierarchical approach and regression approach were used to correct for unequal cell size, while education level and heroin use remained the covariates. CDC groups did not differ in level of perfor-
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
189
TABLE 3 Frequency of Impaired' Test Performance as a Function of HIV Status
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mem I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
HIVf ( N
=
45)
0% (O)*
0% (0) 4% (2) 7% (3) 4% ( 2 ) 4% (2) 0% (0) 9% (4) 4% (2) 13% (6) 18% (8) 7% (3) 20% (9) 13% ( 6 )
HIV- ( N
=
55)
0% (0) 0% (0) 5% (3) 4% (2) 0% (0) 0% (0) 0% (0) 0% (0) 2% (1) 0% (0) 5% (3) 4% (2) 13% (7) 9% ( 5 )
'Performance of two SD below the mean or lower 2percentage (N)
mance on any measure except for the copying trial of the Rey-Osterreith Complex Figure, on which participants with PGL performed significantly worse than the other two groups. T-test comparisons of seronegatives and seropositive asymptomatics (CDC group 11) on all neuropsychological measures yielded no significant differences. The percentage of seropositives and seronegatives who scored in the impaired range for each measure was calculated. When normative means and standard deviations for a test were available, performance of two standard deviations below the mean or lower was considered to be impaired. Cut-off scores for each instrument are described in the method section. The results are shown in Table 3. The percentage of seropositives in each CDC group who scored in the impaired range for each measure was also calculated. These results are shown in Table 4. Finally, the percentages of neuropsychologically impaired participants in seropositive and seronegative groups, and also the three CDC groups were calculated. Impairment is defined as having the scores on two or more measures in the impaired range. The results are shown in Table 5. Chi-square analyses using Yates' correction indicated a highly significant difference in the frequency of impairment between seronegatives and seropositives (? = 22.4, df = 1, p < .001). Chi-square analyses of the frequency of impairment among the three CDC groups was significant (x2= 7.2, df = 2 , p < .05). DISCUSSION These analyses show that pre-AIDS seropositive IVDU in recovery appear to have neuropsychological deficits in the areas of divided attention, visual short-term memory, graphomotor speed and accuracy, auditory contextual language short-term memory, and abstract concept formation. The difficulties in graphomotor speed and accuracy are consistent with the common clinical finding of handwriting deterioration in people with HIV Spectrum Disease (e.g., Forstein, 1989). These deficits cannot be accounted for by age, sex, education, incidence of prior head injury, length of
M.C. WELLMAN
190
TABLE 4 Frequency of Impaired' Test Performance as a Function of Stage of Infection STAGE OF INFECTION
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
NP TEST Pegboard DH Pegboard NDH Logical Mem I Logical Mem I1 Digit Span Vocabulary Block Design Similarities Digit Symbol Trails A Trails B Rey 1 Rey 2 Rey 3
CDC I1 ( N
=
8)
CDC 111 (N
0%(0)2 O%(O) O%(O)
=
O%(O) O%(O) O%(O) O%(O) 9%(1) O%(O) O%(O) 9%( 1) O%(O) 18%(2) 9%(1) 18%(2) 369'44) 9%(1)
O%(O) O%(O) O%(O) O%(O) O%(O) O%(O) 13%(1) 25%(2) 13%(1) 25%(2) O%(O)
11)
CDC IVa ( N
=
26)
O%(O) O%(O) 8%(2) 12%(3) 4%(1) 8%(2) O%(O) 12%(3) 8%(2) 12%(3) 19%(5) O%(O) 12%(3) 19%(5)
'Performance of two SD below the mean or lower 'percentage (N)
drug use, types of drugs used, overdose history, length of time in recovery, use of methadone, and use of AZT or other medication. Van Gorp et al. (1989) reported strikingly similar findings in gay men with fullblown AIDS. These men had significantly worse performance in WAIS-R Digit Symbol, Similarities, and Block Design, Trails A and B, Rey Ostemeth Delayed Recall, and RAVLT Delayed Recognition than did seronegative men. It is interesting to note that in the van Gorp et al. study, only people with AIDS, rather than ARC, significantly differed from seronegatives, whereas in the present study seropositives in pre-AIDS groups significantly differed from seronegatives. Similarly, Tross, Price, Navia, Thaler, Gold, Hirsch, and Sidtis (1988) found that gay men in the early stages of AIDS performed significantly worse than did seronegative gay men on WAIS-R Digit Symbol and Block Design, grooved pegboard, TABLE 5 Neuropsychological Impairment* as a Function of HIV Status and Stage of Infection
HIV( N = 55) HIV + ( N = 45) CDC I1 ( N = 8) CDC 111
Impaired
Unimpaired
11% (6)'
89% (49)
22% (10)
78% (35)
13% (1)
87% (7)
18% (2)
82% (9)
27% (7)
73% (19)
(N = 11)
CDC 1Va (N = 26)
*Neuropsychological impairment is herein defined as performance in the impaired range (2 SD below the mean or lower) on two or more neuropsychological measures. 'percentage (N)
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
191
and Trailmaking A and B, while seronegatives and asymptomatic seropositives did not differ on these or any of the other 12 measures. Some studies report that both heroin and cocaine cause immunosuppression, and over 90% of the subjects in the present study used IV heroin and cocaine in combination. In addition, there was a high incidence of head injury among the IVDU in this study. Both of these factors may contribute to the incidence of neuropsychological dysfunction among the seronegative participants. However, seropositive and seronegatives did not differ significantly in these pre-existing factors, and the seropositives had a significantly greater incidence of impairment than did the seronegatives. Thus, it is possible that once IVDU become infected with HIV, subcortical structures in the brain, which have been implicated as HIV targets in early stages of infection, may become differentially more vulnerable to invading viruses at the time of initial acute infection (CDC Group I). In fact, recent studies by Daar, Moudgil, Meyer, and Ho (1991), and Clark, Saag, Decker, Campbell-Hill, Roberson, Veldkamp, Kappes, Hahn, and Shaw (1991) report that at the time of initial infection, a large explosion of virus replication occurs. The latter researchers found that one milliliter of blood from a newly infected individual may contain 1000 tissue-cultureinfective doses, a level that then drops dramatically when attacked by the body’s Tcells. The data in the present research suggest that there may be an initial “vulnerability factor” (van Gorp, 1991) caused by the pre-existing factors of history of head injury and chronic drug use which cause weakened subcortical structures to be invaded by HIV when the initial explosion of viral replication occurs at CDC stage I. The findings of the present study are also in contrast to studies of gay men which report that the percentage of impaired participants does not increase significantly until participants with AIDS are entered into the equation. For example, Selnes, Miller, Becker, Cohen, McArthur, Visscher, Gordon, Satz, Ginzburg and Polk (1988), using the MACS data, reported that when comparing the performance of 786 seronegtive gay men and 757 seropositive (CDC Groups I1 and 111) on eight neuropsychological measures, 13% of seronegatives and 15% of seropositives were impaired (two measures more than 2 SD below the mean). This difference was not significant. Likewise, Miller, Selnes, McArthur, Satz, Becker, Cohen, Sheridan, Machado, van Gorp, and Visscher (1990), also using the MACS data, reported that on ten neuropsychological measures, 769 seronegatives showed 3.9% impairment (same criteria for impairment as above) while 727 seropositives (CDC Groups I1 and 111) showed 5.5% impairment. While these percentages of impairment were substantially lower than those reported in the first paper, again the differences between seropositives and seronegatives was not significant. Tross et al. (1988) also reported no difference in percentages of impaired participants when comparing seropositives and seronegatives on 17 neuropsychological measures. They found 0% of both seropositives and seronegatives had two “abnormal” tests, while 20% of participants in the early stages of AIDS did have two abnormal tests. In the present study, 18% of IVDU participants with PGL (CDC group 111) had two or more abnormal tests, and 27% of those with constitutional disease (CDC group IVa) were similarly impaired. Finally, Wilkie, Eisendorfer, Morgan, Lowenstein, and Szapocznik ( I 990) reported in their narrative that when comparing 46 seropositive (CDC Groups I1 and 111) gay men with 13 seronegative gay men using 12 traditional neuropsychological measures, 0% of the seronegatives and 3% of the seropositives were impaired (two measures more than 2 SD below the mean), while their Table 2 shows 0% of both groups impaired. In either case, the differences were not significant, even though
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
I92
M.C. WELLMAN
the authors did find specific significant differences between seropositives and seronegatives in four of the twelve traditional neuropsychological measures. Thus, these data suggest that neuropsychological impairment appears earlier in the course of HIV infection among lVDU than it does among gay men. When comparing the scores of the seronegatives and the seropositive asymptomatics (CDC Group 11) on the neuropsychological measures, there were no significant differences in any of the measures. However, as infection progresses to stages of PGL and constitutional disease, impairment was noted in auditory short-term memory for contextual language, logical categorical thinking, graphomotor speed and visual short-term memory. Such deficits are seen frequently in gay men only at the stage of full-blown AlDS. It is hypothesized that this escalation of neuropsychological impairment in IVDU may occur as a result of the initial vulnerability of subcortical brain tissue in these IVDU. Gay men in a number of studies showed few if any signs of neuropsychological impairment at pre-A1DS stages, probably because the gay men in most studies had no history of head injury or drug use, and thus lacked the hypothesized vulnerability factor. Another important point needs to be made concerning the 11% rate of impairment among the seronegatives. While all seronegatives had received their last HIV negative test results less than 12 months prior to the study, two of the six impaired seronegative participants reported three symptoms of constitutional disease upon interview, and two of the six reported four symptoms of constitutional disease. Thus, it is possible that these four participants’ last HIV test failed to detect HIV infection and falsely indicated seronegativity. Grant and Heaton (1990) voiced their concerns about the validity of HIV test results when selecting subjects for research, stating that some people may test negative when in fact they are positive. It is common knowledge that false negatives exist, as the ELISA and Western Blot tests only identify the presence of antibodies, rather than HIV itself. Other research studies have not reported any investigation of possible symptomatology in their seronegative subjects, so this problem of false negatives used in research may be fairly widespread. In this study, if the four out of six impaired seronegative participants who reported three or four constitutional symptoms truly are seropositive, then the actual rate of impairment among seronegatives is 2/51, or 4%, rather than 6/55, or 11%. Thus, a comparison between the percentage of impaired lVDU seronegatives and seropositives, even the asymptomatic seropositives (CDC Group Il), would yield quite different results, and would provide even more strong support for the notion that HIV affects neuropsychological functioning in IVDU early in the course of infection.
REFERENCES Clark, S . , Saag, M . , Decker, W . , Campbell-Hill, S . , Roberson, J . , Veldkamp, P . , Kappes, J . , Hahn, B., & Shaw, G . (1991). High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-I infection. The New England Journal of Medicine, 324, 954-960. Daar, E., Moudgil, T., Meyer, R . , & Ho, D. (1991). Transient high levels of viremia in patients with primary human immunodeficiency virus type 1 infection. The New England Journal of Medicine, 324, 961-964. Donahoe, R. & Falek, A. (1988). Neurc)immunomodulation by opiates and other drugs of abuse: Relationship to HIV infection and AIDS. In T. P. Bridge (Ed.) Psychological, neuropsychiafric and substance abuse aspects of AIDS. New York: Raven Press. Forstein, M . (1989, October). Neurological aspects of HlV infection. Colloquium presented at University of Massachusetts Medical Center, Worcester, MA.
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
HIV INFECTION AND NEUROPSYCHOLOGICAL IMPAIRMENT
193
Goethe, K., Mitchell, J., Marshall, D., Brey, R., Cahill, W . , Leger, G., Hoy, L., & Baswell, R. (1989). Neuropsychological and neurological function of human immunodeficiency virus seropositive asymptomatic individuals. Archives of Neurology, 4 6 , 129- 133. Grant, I . , Atkinson, J., Hesselink, J., Kennedy, C., Richman, D., Spector, S., & McCutchan, J. (1987). Evidence of early central nervous system involvement in AIDS and other HIV infections. Annals of Internal Medicine, 107, 828-836. Grant, I . & Heaton, R. (1990). Human immunodeficiency virus-type 1 (HIV-I) and the brain. Journal of Consulting and Clinical Psychology, 58, 22-30. Levy, R., Janssen, R., Bush, T., & Rosenblum, M. (1988). Neuroepidemiology of acquired immunodeficiency syndrome. Journal of Acquired Immune Deficiency Syndromes, 1, 3 1-40, Lewis, B. Personal communication, August 17, 1989. Lezak, M. (1983). Neuropsychological assessment (2nd ed). New York Oxford University Press. Lunn, S., Skydsbjerg, M., Schulsinger, H., Pamas, J., Pedersen, C., & Mathiesen, L. (1991). A preliminary report on the neuropsychologic sequelae of human immunodeficiency virus. Archives of General Psychiatry, 4 8 , 139-142. Miller, E . , Selnes. O., McArthur, J., Satz, P., Becker, J . , Cohen, B., Sheridan, K., Machado, A , , van Gorp, W., & Visscher, B. (1990). Neuropsychological performance in HIV-1-infected homosexual men: The Multicenter AIDS Cohort Study. Neurology, 4 0 , 197-203. Navia, B. & Price, R. (1987). The AIDS Dementia Complex as the presenting or sole manifestation of HIV Infection. Archives of Neurology, 4 4 , 65-69. 0110, C., Johnson, R., & Grafman, J. (1991). Signs of cognitive change in HIV disease: An eventrelated brain potential study. Neurology, 4 1 , 209-215. Penk, W., Brown, A,, Roberts, W., Dolan, M., Atkins, H., & Robinowitz, R. (1981). Visual memory for Black and White male heroin and non-heroin drug users. Journal of Abnormal Psychology, 90, 486-489. Price, R.,Navia, B., & Cho, E. (1986). AIDS encephalopathy. Neurologic Clinics, 4 , 285-301. Reitan, R. (1986). Trail Making Test: Manual for administration and scoring. Tucson, AZ: Reitan Neuropsychological Laboratory. Rey, A. (1941). L’examen psychologique dans les cas d’encephalopathie traumatique. Archives de Psychologie, 28, 286-340. Selnes, O., Miller, E., Becker, J., Cohen, B., McArthur, J . , Visscher, B., Gordon, B., Satz, P., Ginzburg, H., Polk, B. (1988, June). Normal neuropsychological performance in healthy HIV-1 infected men: The Multicenter AIDS Cohort Study (MACS). Paper presented at the IVth International Conference on AIDS, Stockholm, Sweden, June 15-16, 1988. Selnes, O., Miller, E., McArthur, J., Gordon, B., Munoz, A , , Sheridan, K., Fox, R., & Saah, A. (1990). HIV-1 infection: No evidence of cognitive decline during the asymptomatic stages. Neurology, 4 0 , 204-208. Silberstein, C., McKegney, F. P., O’Dowd, M., Selwyn, P., Schoenbaum, E., Drucker, E., Feiner, C . , Cox, C., Friedland, G. (1987). A prospective longitudinal study of neuropsychological and psychosocial factors in asymptomatic individuals at nsk for HTLV-III/LAV infection in a methadone program: Preliminary findings. International Journal of Neuroscience, 32, 669-676. Stem, Y., Marder, K., Bell, K., Chen, J., Dooneief, G., Goldstein, S., Mindry, D., Richards, M., Sano, M., Williams, J., Gorman, J . , Ehrhardt, A , , Mayeux, R. (1991). Multidisciplinary baseline assessment of homosexual men with and without human immunodeficiency virus: 111. Neurologic and neuropsychological findings. Archives of General Psychiatry, 4 8 . 13 1 - 138. Strang, J . & Gurling, H. (1989). Computerized tomography and neuropsychological assessment in longterm high dose heroin addicts. British Journal of the Addictions, 84. 1011-1019. Thyss, A,, Bessi, C., & Kretzschmar, J . (1982). Cellular immunity in heroin addicts. Psychologie Medicale, 14, 1731-1733. Trites, R. (1987). Instructions for the 32025 Grooved Pegboard Test. Lafayette, IN: Lafayette Instrument Company. Tross, S., Price, R., Navia, B., Thaler, H., Gold, J., Hirsch, D., & Sidtis, I. (1988). Neuropsychological characterization of the AIDS Dementia Complex: A preliminary report. AIDS,2 . 81-88. van Gorp, W. Personal communication, January 24, 1991. van Gorp, W., Miller, E., Satz, P., & Visscher, B. (1989). Neuropsychological performance in HIV1 immunocompromised patients: A preliminary report. Journal of Clinical and Experimental Neuropsychology, 11, 763-773. Wechsler, D. (198 1). Wechsler Adult Intelligence Scale-Revised. New York: The Psychological Corporation. Wechsler, D. (1987). Wechsler Memory Scale-Revised. San Antonio, TX: The Psychological Corporation.
I94
M.C. WELLMAN
Downloaded by [University of California, San Diego] at 00:05 17 March 2016
Wellman, M. C . (1991). Psychological distress among intravenous drug users with HIV infection. Manuscript submitted for publication. Wellman, M . & Allen, M . (1983). Variations in hand position, cerebral lateralization, and reading ability among right-handed children. Brain and Language. 18, 217-292. Wilkie, F . , Eisdorfer, C . , Morgan, R . , Lowenstein, D . , Szapocznik, J . (1990). Cognition in early human immunodeficiency virus infection. Archives of Neurology, 47, 433-440.
